Characterising risk of non-steroidal anti-inflammatory drug-related acute kidney injury: a retrospective cohort study

BACKGROUND: Non-steroidal anti-inflammatory drugs (NSAIDs) are commonly prescribed for pain and inflammation. NSAID complications include acute kidney injury (AKI), causing burden to patients and health services through increased morbidity, mortality, and hospital admissions. AIM: To measure the extent of NSAID prescribing in an adult population, the degree to which patients with potential higher risk of AKI were exposed to NSAIDs, and to quantify their risk of AKI. DESIGN & SETTING: Retrospective 2-year closed-cohort study. METHOD: A retrospective cohort of adults was identified from a pseudonymised electronic primary care database in Hampshire, UK. The cohort had clinical information, prescribing data, and complete GP- and hospital-ordered biochemistry data. NSAID exposure (minimum one prescription in a 2-month period) was categorised as never, intermittent, and continuous, and first AKI using the national AKI e-alert algorithm. Descriptive statistics and logistic regression were used to explore NSAID prescribing patterns and AKI risk. RESULTS: The baseline population was 702 265. NSAID prescription fell from 19 364 (2.8%) to 16 251 (2.4%) over 2 years. NSAID prescribing was positively associated with older age, female sex, greater socioeconomic deprivation, and certain comorbidities (diabetes, hypertension, osteoarthritis, and rheumatoid arthritis) and negatively with cardiovascular disease (CVD) and heart failure. Among those prescribed NSAIDs, AKI was associated with older age, greater deprivation, chronic kidney disease (CKD), CVD, heart failure, diabetes, and hypertension. CONCLUSION: Despite generally good prescribing practice, NSAID prescribing was identified in some people at higher risk of AKI (for example, patients with CKD and older) for whom medication review and NSAID deprescribing should be considered.